Prototype, ready for a utility pilot

JalaLens shows a water board where its water is being lost.

AI decision support for city water boards. JalaLens joins pipe maps, flow data, billing and building footprints into one daily, ranked field-visit list per sub-division: likely leaks, buildings near a main with no account, and suspicious bills. Each lead says why.

See how JalaLens works Email about a pilot
Illustration: one sub-division drawn as a grid of hexagons over its water mains, with three hexagons ranked for a field visit 1 2 3
  1. 1
    Likely leak area Night flow up for 6 nights, many old valves
  2. 2
    Building, no account Within 30 m of a main, no billed connection
  3. 3
    Suspicious bill Flat reads for months on a domestic tariff
SYNTHETIC Illustration with made-up data, not a real utility.

A large share of city water earns nothing

Bengaluru shows the scale. The board already runs a SCADA centre, field teams and data matching. What is missing is one tool that tells each sub-division where to send its teams next.

How it works, in three steps

  1. Connect the data you already have

    Pipe and valve maps, bulk-flow exports from SCADA and a billing extract. JalaLens installs on one server inside the utility, so the data stays there.

  2. Find and rank the leads

    Water balance, night flow and anomaly detection find likely leaks. Map joins find buildings near a main with no account. A scorer flags odd bills.

  3. Send teams, then learn

    A daily list sized to team capacity, as a map, CSV or API. A lead means “inspect”; an officer decides. Field results train the next list.

Built for city water boards

JalaLens sits on top of the systems a board already runs. It reads their outputs and adds the ranking, so it complements robots, SCADA and field teams rather than replacing them.

  • Sub-division engineersA short list of where to look today, with the reason and the litres or rupees at stake.
  • Revenue and vigilance teamsLikely unbilled connections and billing anomalies, checked by an officer before any action.
  • Command centres and GIS desksResults as GeoJSON, CSV and a documented REST API, loaded as one more map layer.

Also from Kandukuri AgriGroup

JalaLens is our lead product. The company also runs these lines.

Banana exportsG9 Cavendish bananas from Andhra Pradesh, shipped from Chennai. Trade deskImports into India through our IEC registration. Aerospace R&DResearch into a hybrid-petrol UAV with on-board, offline computing.

Run a 60-day pilot with us

Two sub-divisions, your own data, measured against random field visits.

business@agrigroup.online